Iteratively applying multiple ROIs for subject specific network analysis from ommand-line

[William McCuddy]

Hello, 

Does anyone have an example command for iteratively applying pairs of ROIs from a list of 17 for a total 136 ROI-ROI contingencies. The 17 ROIs have been selected from the HCP-MP Atlas based on their overlap with the default mode network. The goal is to automatically generate tracts comprising the DMN. I might also need to add in some ROAs to keep every tidy, but I can probably figure that out later.    

These are the specific ROIs

1 L_10r

2 L_a24

3 L_p32

4 L_s32

5 L_31a

6 L_31pd

7 L_31pv

8 L_d23ab

9 L_POS1

10 L_POS2

11 L_RSC

12 L_v23ab

13 L_IP1

14 L_PFm

15 L_PGi

16 L_PGs

17 L_TPOJ3 

Thanks!

Travis

[Frank Yeh]

First make sure you have 17 regions saved as one nii.gz file (DSI
Studio should also have its region labels as nii.gz.txt)
Then the following command should work out for you:

dsi_studio --action=trk --source=subject.fib.gz --fiber_count=1000000
--output=no_file --connectivity=my_HCP_MMP.nii.gz
--connectivity_value=trk --connectivity_type=pass

Best regards,
Frank

> --
> You received this message because you are subscribed to the Google Groups "DSI Studio" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to dsi-studio+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/57fea5bb-01d9-420f-8fb4-77286bfbae9cn%40googlegroups.com.

[William McCuddy]

Thanks for quick response! I tried running the cmd in a bash script (see below) and replaced [mysubject] and [my_HCP_MMP] with my specific file names. The script runs, but i get all 1000000 streamlines when i open the resulting .trk file in the gui. I made sure to have all 17 regions saved as one nii.gz file (which i did via Regions-->Save all regions as...  I'm able to open this file in the gui and see all 17 ROIs displayed). The nii.gz.txt appears to be fine as well. I wasn't sure if these files needed to be save in the dsi-studio folder or the subject folder (where the .fib file is saved). I tried saving copies in both places and got the same result. I also tried to make another nii.gz file with just two regions and I got the same result (.trk file showing all 100000 streamlines). 

This is what the simple script looks like:

path=D:\Imaging dsi_studio64

dsi_studio --action=trk --source=[mysubject].fib.gz --fiber_count=1000000

--output=no_file --connectivity=[my_HCP_MMP].nii.gz
--connectivity_value=trk --connectivity_type=pass

Any idea what I'm doing wrong?

[Frank Yeh]

It must be a problem in DSI Studio.
Could you add > log.txt at the end of your command and post the
content of log.txt output?
I will fix it as soon as possible.
Frank

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/bd99f795-9f78-41ed-9d93-8e686283ea7dn%40googlegroups.com.

[William McCuddy]

Here is output of cmd line (changed SubjectFib to deidentify):

D:\DICOMOBJ\[SubjectFib]>--output=no_file --connectivity=L_DMN.nii.gz

'--output' is not recognized as an internal or external command,

operable program or batch file.

D:\DICOMOBJ\ [SubjectFib] >--connectivity_value=trk --connectivity_type=pass

'--connectivity_value' is not recognized as an internal or external command,

operable program or batch file.

D:\[SubjectFib]>log.txt

Here is output of text file:

DSI Studio Jul  1 2020, Fang-Cheng Yeh

source=sample.fib.gz

action=trk

loading sample.fib.gz...

sample.fib.gz does not exist. terminating...

It feels like I'm doing something wrong (likely), but DSI knows SubjectFib.gz exists because i get a SubjectFib.gz.trk.gz output.

Thanks for your help!

[Frank Yeh]

Could you update DSI studio and try again?

Frank

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/5c9ab805-487e-4b2b-b309-64d8797eb8e5n%40googlegroups.com.

[William McCuddy]

Stupid question. To update, do i need to download latest version (and if so, is this the appropriate link for download:http://dsi-studio.labsolver.org/dsi-studio-download) or is there a way to update from the cmd line?

Thanks

[Frank Yeh]

Sorry it cannot be updated from the command line. You would need to
get it from the download link you mentioned.
Frank

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/ceaa02f1-5a3a-4b43-a40e-fbd850e4f323n%40googlegroups.com.

[William McCuddy]

Updated (GUI says "June 2 2021 build")

I generated new src file and fib file, re-ran command, and got same result except now I get .tt.gz file 

Output of log.txt:

D:\[path to subject folder containing new .fib] >path=D:\Imaging\dsi_studio_64

D:\[path to subject folder containing new .fib]  >dsi_studio --action=trk --source=D:\ [path to subject folder containing new .fib]\my subject.fib.gz --fiber_count=1000000

D:\[path to subject folder containing new .fib]  >--output=no_file --connectivity=L_DMN.nii.gz

'--output' is not recognized as an internal or external command,

operable program or batch file.

D:\ [path to subject folder containing new .fib]  >--connectivity_value=trk --connectivity_type=pass

'--connectivity_value' is not recognized as an internal or external command,

operable program or batch file.

D:\ [path to subject folder containing new .fib]  >log.txt

----------

both L_DMN.nii.gz and L_DMN.nii.gz.txt are located in the main DSI studio folder and the subject folder containing fib.gz and .bat file. 

[Frank Yeh]

You have to put all commands in the same line...(except for the first
path= xxxx)
Frank

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/8771ab8f-2274-4ffc-9033-8213a91f3bden%40googlegroups.com.

[William McCuddy]

oh, duh! That worked. Thank you helping me figure this out. It is now creating a .tt.gz file for each region ROI. Will it automatically generate a merged file? if not could i add another cmd line to the .bat file? 

The closest thing i can find in online documents is:

cluster: run track clustering after fiber tracking. 4 parameters separated by comma are needed: --cluster=METHOD_ID,CLUSTER_COUNT,RESOLUTION,OUTPUT FILENAME

 

                  METHOD_ID: 0=single-linkage clustering 1=k means 2=EM

                  CLUSTER_COUNT: The total number of clusters. In k-means or EM, this is the total number of clusters assigned. In single-linkage, it is the maximum number of clusters allowed to avoid over-segmentation (the remaining small clusters will be group in the last clusters).

                  RESOLUTION: only used in single-linkage clustering. The value is the mini meter resolution for merging the clusters.

[Frank Yeh]

No, it doesn't. You can merge them using GUI.
I am sorry that the command line does not have a track merging function.
Frank

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/5388fea3-cd8e-4d25-8cdc-5dc57381db58n%40googlegroups.com.

[William McCuddy]

Looks like I spoke too soon. While it did create individual tract files for each ROI pair, when opening each tract file there was still over 100000 streamlines for every single pair (instead of showing only those tracts that pass through the two ROIs for each pair). Any idea why?

[William McCuddy]

this is the command, all in single line (except for the first line with Path to dsi studio, which is not shown)

dsi_studio --action=trk --source=D:\DICOMOBJ\Sub001\401_DTI_32dir_3iso\Sub001\Sub001.src.gz.odf.gqi.1.25.fib.gz --fiber_count=1000000 --output=no_file --connectivity=L_DMN.nii.gz --connectivity_value=trk --connectivity_type=pass > log.txt

[Frank Yeh]

The track numbers aren't necessarily wrong.
You may open the tt.gz file in Step T3 Fiber tracking to see if the
tract did pass through the designated ROI.

Best regards,
Frank

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/7af103ab-35a4-47b9-b39f-6510388c764en%40googlegroups.com.

[William McCuddy]

Thanks for help and your patience, still just trying to figure this thing out.  

Below is pic of 2 rois with manual tracking (left) and the same 2 rois tracted with cmd line (right). The picture on the right is how all tracts generated from paired ROIs in cmd line look. 

[Frank Yeh]

Could you post the log.txt here?

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/9de983e6-3459-4fce-aae9-dc087cb1bbbbn%40googlegroups.com.

[William McCuddy]

It is giving me the same errors as before. I edited the directory names to remove subject/patient names so i could post exact log.txt ([OldDirectoryName]=Sub001). However, even though i also edited the path in .bat file (below), the log.txt file still shows old directory names. I also changed [OldFileName]...fib.gz to Sub001...fib.gz

Cmd line from .bat file (minus first line/path to dsi_studio): 

dsi_studio --action=trk --source=D:\DICOMOBJ\Sub001\401_DTI_32dir_3iso\Sub001\Sub001.src.gz.odf.gqi.1.25.fib.gz --fiber_count=1000000 --output=no_file --connectivity=L_DMN.nii.gz --connectivity_value=trk --connectivity_type=pass > log.txt

log.txt output:

D:\DICOMOBJ\[OldDirectoryName]\401_DTI_32dir_3iso\[OldDirectoryName]>path=D:\Imaging\dsi_studio_64

D:\DICOMOBJ\[OldDirectoryName]\401_DTI_32dir_3iso\[OldDirectoryName]>dsi_studio --action=trk --source=D:\DICOMOBJ\[OldDirectoryName]\401_DTI_32dir_3iso\ [OldDirectoryName]\[OldFileName].fib.gz --fiber_count=1000000

D:\DICOMOBJ\[OldDirectoryName]\401_DTI_32dir_3iso\[OldDirectoryName]>--output=no_file --connectivity=L_DMN.nii.gz

'--output' is not recognized as an internal or external command,

operable program or batch file.

D:\DICOMOBJ\[OldDirectoryName]\401_DTI_32dir_3iso\[OldDirectoryName]>--connectivity_value=trk --connectivity_type=pass

'--connectivity_value' is not recognized as an internal or external command,

operable program or batch file.

D:\DICOMOBJ\[OldDirectoryName]\401_DTI_32dir_3iso\[OldDirectoryName]>log.txt

[Frank Yeh]

You will need to make sure that all parameters are in the same line of the command.

There seems to be a line break that cuased the problem.

Frank

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/083ccab6-e705-4c98-833a-1042cb12ec87n%40googlegroups.com.

[Frank Yeh]

You may need to use a different text editor that surely won't automatically break a line into multiple lines.

Also, check log.txt output.

Otherwise, your command line won't work correctly.

Frank

[William McCuddy]

confirmed text is all in one line by expanding window (pic below). still same error. I'm going to start from scratch in new directory to see if issue is reproduced. 

[William McCuddy]

OK, starting from scratch at least gave me a log.txt file that makes more sense (below), but same issue as before (one ROI pair tt.gz files contains tracts of whole brain instead of only the tracts that pass through the 2 rois; as seen in pic) Note. I abbreviated the region and label file (L_DMN_test) to contain only 4 regions while testing.

log.txt output:

 DSI Studio Jun  2 2021, Fang-Cheng Yeh

source=D:\DICOMOBJ\Test\Sub001.src.gz.odf.gqi.1.25.fib.gz

action=trk

loading D:\DICOMOBJ\Test\Sub001.src.gz.odf.gqi.1.25.fib.gz...

otsu_threshold=0.6

fa_threshold=0

dt_threshold=0.2

turning_angle=0

step_size=0

smoothing=0

min_length=30

max_length=300

method=0

initial_dir=0

interpolation=0

seed_plan=0

random_seed=0

check_ending=0

tip_iteration=0

fiber_count=1000000

seed_count=0

thread_count=4

start tracking.

thread_count=4

finished tracking.

1000000 tracts are generated using 2109424 seeds.

0 tracts are removed by pruning.

The final analysis results in 1000000 tracts.

output=no_file

connectivity=L_DMN_test.nii.gz

connectivity_type=pass

connectivity_value=trk

loading L_DMN_test.nii.gz

DWI dimension=(128,128,55)

NIFTI dimension=(128,128,55)

nifti loaded as multiple ROI file

looking for region label file

label file loaded:D:/DICOMOBJ/Test/L_DMN_test.txt

a total of 4 regions are loaded.

connectivity_threshold=0.001

count tracks by passing

calculate matrix using trk

saving L_DMN_1_L_DMN_2.tt.gz

saving L_DMN_1_L_DMN_3.tt.gz

saving L_DMN_1_L_DMN_4.tt.gz

saving L_DMN_2_L_DMN_3.tt.gz

saving L_DMN_2_L_DMN_4.tt.gz

saving L_DMN_3_L_DMN_4.tt.gz

[William McCuddy]

I think I am getting close to figuring this out. The original command you gave me did not have --roi parameter.  In the cmd line I added "-roi=L_DMN_test.nii.gz" between --fiber and --output. 

Looking at online documentation the following is provided: "(e.g. --roi=nifti.gz:1 only load regions with value=1)" What does the "1 only load regions with value=1" mean?

[Frank Yeh]

Thank you for your help.

I finally found the bug and fixed it.

Please update DSI Studio again (dated today, 1:43 pm) and the updated version should generate a much smaller tt.gz file.

If not, please let me know.

Sorry for the hassle!

Frank

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/8151e812-cb7b-45bf-86f3-48b3ef750450n%40googlegroups.com.

[William McCuddy]

Thanks Frank, this is definitely better, but it automatically generated tracts between RO1 and ROI2 still significantly outnumber the manually generated tracts between the same two ROIs (see pics below). Are the standard tracking parameters different when running cmd line vs gui?

Also, no hassle at all! This is a really cool program and I am very appreciative of all the time and energy you have put into it! 

[Frank Yeh]

Yes, there are differences.

The default seed/tract count at [Step T3(c) Options][Tracking Parameters][Terminate If] is only 50000, but the one assigned at --fiber_count is 1000000

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/a50841e5-17d6-4d30-aad2-887520902bf6n%40googlegroups.com.